论文部分内容阅读
提出一种基于多尺度空间聚集(MDSC)的犯罪热点探测方法,与蒙特卡罗模拟下的随机分布结果相比,MDSC方法探测的热点具有统计显著性。它以K函数为理论基础,根据不同尺度自动确定犯罪热点探测参数,并可以在其他要素分布基础上动态调整聚集所需最少点数,实现了基于特定基准变量的热点探测风险调整,并可满足对犯罪高发地区和高危地区的热点探测需求,结果具有客观性和可对比性。
This paper proposes a crime hot spot detection method based on multi-scale spatial aggregation (MDSC). Compared with the random distribution results under Monte Carlo simulation, the hot spots detected by MDSC method are statistically significant. Based on the K function, it automatically determines the detection parameters of crime hot spots on different scales and dynamically adjusts the minimum number of points required for aggregation based on the distribution of other factors, and realizes the hot spot detection risk adjustment based on specific reference variables, High-crime areas and high-risk areas to detect the hot demand, the result is objectivity and comparability.